response time - TN vs VA vs IPS

[blexxun]

I've found a research paper comparing the response time of the various LCD display technologies (see source info below).
The paper clearly outlines the advantages and disadvantges of TN, MVA and IPS with respect to grey turn on and off response. The paper says:
"In TN and MVA devices, on-time response including all
grey scales is widely spread in a range of 4..60ms, that is,
the ratio (R) of slowest to fastest response is about 10
but the IPS case is about 2.2 For off response time,
the R values are almost same among three devices, not
very large though the absolute value of the IPS device is
relatively small."

From the measurement results in the paper (its a 3d chart which is a bit hard to read due to bad picture quality)
my conclusion is the following:

grey response time:
IPS is the best
TN is medium and
MVA is worst!

With respect to white-black response (and this data is the only one specified in the response time of LCD monitors ):

IPS & MVA better than TN

The summary of the cited paper is:

"In image quality viewpoint, 2-domain IPS and FFS devices seems best and are comparable to the CRT display though still needs further in color reproduction and low contrast ratio in dark environment."

The additional LC mode which is mentioned in the paper called 'FFS' (=fringe field switching)has similiar good properties like IPS and seems a bit faster. It is already in massproduction (Hunday Display).
I've found the specs of a FFS-panel which has a response time of 20ms (this is of course only the black-white response). You can find the panel data here:
http://www.hydis.com/product/monitor/p_view.asp?p_id=3

The weak point of this panel is the small viewing angle. This is due to the fact that it is only single domain. A dual domain version called 'Ultra-FFS' with a better viewieng angle is already announced:
http://www.hydis.com/technology/ffs_tech/ffs_future.shtml


It seems that the paper I've found confirms the test results, that MVA panels are not so good for 'gaming'. However, it is not clear to me right now if all VA-technologies (thats MVA,PVA,VA-IPS according to that paper) are in the same boat when it comes to response time (but it seems so).
If one is looking for shortest possible response time one has to look for FFS panels. A monitor which such kind of panel is available in May:
http://www.chip.de/news_stories/news_stories_8718454.html
But it has a small viewing angle.

The next best choice are dual domain IPS monitors, such as EIZO L565.

I'll review all monitor reviews again, to find out if there is a tendency that IPS LCD-Monitors are indeed better.


source:


"Liquid Crystal Displays with High Image Quality and Fast Response Time"
S. H. Lee and M-H. Lee
Department of Polymer Science and Technology, Chonbuk National University, Chonju 561-756
Journal of the Korean Physical Society, Vol. 39, December 2001, pp. S42-S48
(Received 1 November 2000, in final form 12 September 2001)

<P ID="edit"><FONT SIZE=-1><EM>Edited by blexxun on 04/14/02 07:21 AM.</EM></FONT></P>

 

[flamethrower205]

Better not have been the research paper I'm writing! j/k Hmm, very useful- that could explain the reason why some are unsatisfied w/ MVA.....

Sig of the week.

 

[Anemone]

In the end Blexxun you went with a PVA Samsung panel did you not? Were you time constrained or do you honestly think in the next few months nothing dramatic is coming?

I ask because from all the press, I'm getting the impression that the latter half of this year and the first quarter of next, there are some fairly impressive things coming that might leave me happy with crt for just a bit longer. Curious anyway.

 

[blexxun]

Right...I replaced the 19" CRT on one of my PCs by a Samsung 171p. For my second PC, however, with a triple CRT setup I haven't found a nice (=cost effective) LCD equivalent. But I'm hesitating to buy another 3.

While the 171p is nice its still far from being perfect and it might be possible that other LCD monitors have a better visual performance.
Due to the lack of qualified monitor tests I have to find out that for myself. This will take some time. At least my idea to evaluate LCD monitors according to their technology TN vs IPS vs (M)VA is exactly the right one. Simply speaking thats similiar to computers which are separated according to their CPUtype Pentium, Celereon, Athlon and Duron.

It is also clear (to me) that fast repsonse time of PVA/MVA for black/white transitions has to be paid by a very slow response for gray response. This was not only pointed out in the cited paper at the beginning of this thread but even by Fujitsu itself. This is the major reason why I'm not jumping on the MVA/PVA train right now.

On last years SID conference Fujitsu presented a paper stating to have improved their MVA technology with respect to gray response.

It is now 30ms or 1/3rd of the old one!

Hear hear..the MVA technology was slow!
MVA is dead.....long live "MVA Premium"!

In this paper Fujitsu compared the new MVA to the old one and to Super IPS. Yes, S_IPS was the reference for grey response! Of course, they are now better than Super-IPS.
In the meantime they've made this new panel technology available for mass production (19in. and bigger). Surprisingly, on their homepage they dont even spend a single word about improved gray response time for "MVA premium"....strange...
At least the conference presentation confirms that the grey response of MVA is slow.

My own tests on the 171P are telling me that the grey response of PVA is slow as well. Not that I'm complaining about my 171p but may be there is something better...

Do I have to go for a Super-IPS monitor because it has a better grey response or is the fast black/white response of MVA/PVA more important? I have no clear answer for that right now. Right now I think a fast black/white response is more important because high contrast changes are recognized most easily.

Sure, thats a very subjective matter.

talking about subjective evaluation..Here's another story.....
A nice paper was presented by Samsung people at the SID last year. Title is:
"LCD's, how Fast is Enough? "
This paper is quite interesing.
First, they identified 3 motion artefacts resulting from slow response:
1. decrease of dynamic contrast
2. stroboscopic motion
3. blurred edge

An experimental study was made where people had to identify these artefacts by visual inspection. The results were as follows:

"dynamic contrast reduction" was readily reecognized by all 27 evaluators.

"stroboscopic motion" was immediately recognized by 10 evaluators, 12 recognized it only after explanantion and 5 did not unterstand the artifact at all. (huh)

"blurred moving edge" artifact were only identified by 6 of the 27 subjects. (wow)

It is interesting to see that in one case pple could see the artifacts only after explanantion and in another case
not even after explanation.

Obviously judging image quality viewing a moving image on a LCD display is a highly subjective matter.

This might be the reason why some ppl are giving their LCD monitor back saying they are not acceptable for games like UT or Quake while others have no problems.

But one thing is clear, the blurred edge artifact will never disappear because of the technological limits and the fact that a LCD display is a sample-and-hold device. As we learned from the Samsung report some ppl don't recognize this artefact while other might notice it only after reading a monitor test. So...be careful!

On the other hand, technologies to minimize motion artefacts are known for a couple of years now. But it seems that introduction to mass production is very slow. Could be due to the high priced panel technology and price erosion for the monitor itself. Since most of the known solutions to make panels faster are not cost neutral they won't come.

 

[Anemone]

Good evening Blexxun

I see you’ve been pondering this a lot. I agree, with your summary point about it taking a long time to get these technologies into production panels, which I suspect is partly due to price and probably partly due to where in the component stream some of them should be implemented. FFD for example is a software/controller style solution while some of the technology has to be built into the panel itself. Certainly when profits evaporated, and panels were oversupplied there was little chance of someone getting a technology change to panel design passed by a budget committee.

The rest of this I have read several times, since you’ve packed in a lot of information. From what I’ve been able to glean, subjective plays in, PC and controller card and their age during the development of the DVI specification, as well as the very likely possibility that a given monitor model could theoretically contain one of several different actual panels all gives rise to a high degree of subjectability to the various user reviews. Unless the professional reviews are deliberately trying to throw us, I am highly suspicious that even given the same model, there is variation. Honestly they only have to have panels that meet the specifications they have published, and there are a lot of things there are no specifications for. There are too many “color tests” or standard “should the be same thing” types of tests that turn up different numbers.

Now if I understand correctly, the 172T and 152T as well as all the T Samsungs are based on MVA or variant similar technology. Yet these seem to hold up to gamer review fairly well. I don’t know exactly how it would be done. Could FFD style technology be implemented gradually? I’m thinking that if you had a particular gray scale area that was weak, high response time, could you target that as the first thing for your FFD controller writers to fix and leave the rest over time? Mind you, no change would have to be made to the model #, since you would just be gradually improving a product stream over time. Anyway the heart of the matter is I am wondering where the Viewsonic mva VG191 got rather bad reviews, and the 191T got rather good ones, I am beginning to think the panel is only a small part of the picture, and what is being built up by the manufacturers is a battle over brand loyalty, by differentiating what would otherwise be nearly identical products.

I do have a question for your research oriented mind. Samsung produces a panel 19” that is new and has a 1600x1200 resolution. Panel searches for supply companies shows that they are currently the only ones to produce this resolution in this size, most other manufacturers are only doing that resolution in 20.1 panels. Can you tell what technology this is based off of? My guess is PVA since the panel specifications seem very close to the 171T panel, but would like a second opinion. Personally a 19” that had the 4:3 ratio as well as Samsung quality would sell like hotcakes unless the price was horrendous, which it might be. Just more food for thought.

I note the various technologies you mention, and I have to laud your next point, how fast do we need? Now while a video card speed of say 60-80fps is nice because we all know there are slowdowns, certainly a display that can maintain a steady 35-45 fps is quite likely plenty. Now I know some will disagree, but let me add to that, that if a display cost difference were 2000-4000 USD higher for a 60fps rate, ie an imaginary 17ms response panel, and only 1 in 1000 customers could tell the difference between it and the cheaper panel, its pretty clear which would be a better option from a sales standpoint. Often we get tied up in specifications, without stopping to ask, is what we see satisfactory? And we cannot stress enough, that if it is, don’t spend more! Get what suits the job because a seller will sell you as much as they can get you to spend.

Anyway I have my eye on that 19” panel when it’s a working product, but I do FPS game, as well as work on large financial model spreadsheets, so I’ll have to weigh out its usefulness in both when it becomes a reality. I am pretty much convinced at this point that there are a number of things that affect quality at this point, and trying them out with a good return policy is perhaps the heart of wisdom.

That and being grateful for those who take the time to write personal experience reviews.

Thank you for your extensive research and relating it to us.


<P ID="edit"><FONT SIZE=-1><EM>Edited by Anemone on 05/27/02 12:12 PM.</EM></FONT></P>

 

[Barbatos]

What is that Samsung 172T model you mentioned, I've never heard of it before.

I did a quick search (pricewatch and google) and didn't have anything come up.

 

[Anemone]

New models to replace 151T and 171T I believe

http://samsungelectronics.com/monitor/lcd/index.html

 

[flamethrower205]

Jeez, 350CD/M^2 brightness?? That's wayyy too high.

My frog asked me for a straw...dunno what happened he's all over the place

 

[Anemone]

No different than Viewsonics new high brightness lcd's. To marketing more is better, but I'm of the mind that given daylight and all, 300ish isn't always out of the question. Anyway there are other areas I would like to see improved as I've written before.

 

[flamethrower205]

The human eye needs 40cd/m^2 brightness to see optimally (that means least eyestrain, etc.). Now, I used to have a TFT7020, and can tell you that when brightness was even at 200 cd/m^2, it hurt my eyes, and after about 15 minutes, there was a very noticeable effect. 300 is out of the question, and the way things are going, 400 will be next!

My frog asked me for a straw...dunno what happened he's all over the place

 

[blexxun]

Hello anemone,
about new technologies you said "...where in the component
stream some of them should be implemented", I think this a
very important point. Its not only the panel itself a lot
depends on the panel controller as well. Unfortunately
there exists an "unusually aggressive pricing environment"
in this market segment (O-tone Prudential securities about
Genesis Microchip, whose stock fell almost 30% today
because of this comment; Genesis makes panel controller
ICs)

Also, we currently see a trend to make panels bigger ...not
faster. LCD-TV is the driving force here.

You made another point saying that new technolgies could be
implemented gradually. Indeed, some specifcations of LCD
panels make me think that acceleration techniques are
already there. But not to make the best panels faster.
Instead, they are implemented to make cheap technologies
(like TN) competitive.


About the 19" Samsung panel...yes, its PVA . A short data
sheet about this panel, LTM190U1, can be found at Samsungs
semicondutor site:

http://www.samsungelectronics.com/semiconductors/TFT_LCD/Monitors/LTM190U1/ltm190u1.htm


The question of how much speed we actually need is quite
difficult to answer.
Let me cite from a white paper written by Dave Marsh (Microsoft) about "temporal rate conversion".
There is a chapter entitled "Use flat-panel displays
instead of CRTs " (link is here:

http://www.microsoft.com/hwdev/archive/TVbroadcast/temprate.asp)

He says:

"Personally, I would prefer to watch a flat-panel display
with smearing, rather than have to watch a CRT with flicker
or a CRT with judder."

But then he added:

"My preferred choice, though, would be a CRT display that
did not have flicker or judder because it produces the best
image quality"

There are other articles dealing with this question:

"30 Frames per Second vs. 60 Frames per Second" by Joshua Walrath

http://www.penstarsys.com/editor/30v60/30v60p3.htm

All this is related to your remark, "is what we see satisfactory?"

On one side we have the technical data and on the other
side our visual perception. So far only profesional monitor
tests take care of this, the DICOM standard for medical
applications is an example.


Right now I'm trying to identify all aspects which are
relevant to pick the best LCD monitor.


I'm just at the beginning...

 

[flamethrower205]

Hehe, well that's why it's important to see the monitor yourself.

My frog asked me for a straw...dunno what happened he's all over the place